λ ≈ 1.15 excess air on biomass
Combustion trim target
Boiler · Measurement & control
Boiler instrumentation — the measurement set that protects a boiler
Boiler instrumentation earns its cost with a short list of measurements: drum level under three-element control, steam and feedwater flow, furnace draught, O₂/CO trim toward λ ≈ 1.15, and stack temperature — where 20 °C of reduction is worth about 1 percentage point of efficiency. Arrow Energy Co., Ltd. supplies and integrates this set during boiler and ESP retrofits.
+0.7–0.8 percentage points
Air-trim efficiency value
≈ 20 °C reduction ≈ +1 pp efficiency
Stack temperature rule
3-element: level + steam flow + feed flow
Drum level control
01 — The set
The short list that actually protects a boiler
Most boiler damage is measurable hours or months before it happens — by about a dozen instruments.
A biomass boiler carries hundreds of instrument tags, but its protection rests on a short list: drum level, steam and feedwater flow, furnace draught, flue-gas O₂ and CO, stack temperature, and — where an ESP or bag filter follows — stack opacity and the collector's own electrical readings. Everything else is diagnosis; these are protection and economics. On the boiler products pages the recurring theme is that hardware fails at known temperatures, levels and velocities; instrumentation is how the plant sees those variables before the metal does.
| Measurement | Typical instrument | What it protects / earns | Working figure |
|---|---|---|---|
| Drum level | 2–3 × DP transmitters + gauge glass | Tubes from low-water overheat; turbine from carry-over | 3-element control |
| Steam / feedwater flow | Nozzle or orifice + DP, density-compensated | Level feedforward; efficiency accounting | ±1–2 % of reading |
| Furnace draught | Draught-range DP transmitter | Casing and personnel; combustion stability | −0.05 to −0.2 mbar |
| Flue-gas O₂ / CO | Zirconia probe + IR CO analyser | Stack loss; unburnt loss and fouling | λ ≈ 1.15 target |
| Stack temperature | RTD/thermocouple after last surface | Fouling detection; efficiency | 20 °C ≈ 1 pp |
| Stack opacity | Transmissometer | Collector health; compliance early warning | continuous, % |
| ESP fields | kV / mA per T-R set, spark rate | Field-by-field collection performance | per field |
02 — Drum level
Drum level and three-element control
The one measurement that both trips the boiler and lies to it.
What is three-element drum level control?
A scheme combining drum level, steam flow and feedwater flow. Steam flow is the feedforward — a load change adjusts the feed valve immediately, without waiting for level to move. Feedwater flow closes a fast inner loop around the valve, and measured level trims the residual error. It exists because level alone misleads during load changes.
The lie is shrink and swell. When steam demand jumps, drum pressure dips, steam bubbles below the water surface expand, and indicated level rises — exactly when the drum is actually losing inventory. A single-element controller reads the swell and closes the feed valve, and on a hard swing that ends in a low-level trip. Three-element control ignores the false movement because the flow-balance terms carry the real story. On bagasse boilers, whose fuel feed swings with cane supply minute by minute, the difference between single- and three-element control is the difference between riding through a mill stop and tripping on it. The level measurement itself needs redundancy — two or three DP transmitters with density compensation (a drum at 45 barg holds water at ~257 °C whose density is far from the calibration default), validated against the direct-reading gauge glass, with the gauge glass and its spare glasses and seals maintained as the reference the transmitters are checked against.
03 — Combustion
Draught, O₂/CO trim, and the two cheapest health indicators
Combustion measurement is where instrumentation pays its own invoice.
Furnace draught is held slightly negative — typically −0.05 to −0.2 mbar at the furnace outlet — by the ID fan. Too positive and hot gas and ash puff from every casing opening; too negative and cold air in-leakage dilutes the gas, loads the fan and chills the economizer approach. The draught transmitter is also the first witness of a tube leak: a sustained shift at constant load means the gas-side balance changed.
O₂/CO trim. Biomass units habitually run high excess air because operators add margin against wet-fuel flameout. The cost is stack loss: every kilogram of surplus air is heated from ambient to stack temperature and thrown away. Trimming toward λ ≈ 1.15 — about 3 % O₂ dry — is worth +0.7–0.8 percentage points of boiler efficiency on units running high excess air, with the CO analyser guarding the lower boundary: when CO climbs above a few hundred mg/Nm³, the trim has gone below the combustion-air minimum for the current fuel moisture and must back off. The secondary benefit lands downstream: less excess air means less gas mass, lower fan power, and a lower gas velocity through the ESP — collection efficiency rises with residence time at no capital cost.
Which two instruments tell you the most about boiler health?
Stack temperature and stack opacity. Stack temperature creeping up at constant load means fouling somewhere in the heat-recovery train — about 1 percentage point of efficiency per 20 °C of rise, and it names the problem before an outage inspection can. Opacity rising at constant load flags ESP or bag-filter deterioration days before any stack test.
Both instruments are cheap relative to what they watch. A stack thermometer effectively monitors every heating surface upstream of it — economizer fouling, air-preheater plugging, superheater slagging all raise it, and the soot-blowing schedule can be run against its trend instead of the calendar. Opacity correlates with dust concentration well enough, once calibrated against an isokinetic stack test, to act as the daily proxy between compliance measurements.
04 — ESP side
Instruments on the gas-cleaning train
An ESP announces its problems electrically before the stack shows them.
For plants with an electrostatic precipitator, three signal groups belong on the control-room trend list. Per-field kV and mA from each transformer-rectifier set, with spark rate: each field has a characteristic V–I operating point, and drift diagnoses the machine — falling kV at rising spark rate points to misalignment or a swinging electrode; high kV at near-zero mA points to back corona from high-resistivity dust or a dust-buried electrode. Opacity ties the electrical picture to the outlet. Hopper level switches guard against the most common avoidable ESP trip: a hopper filling until the dust column reaches the internals and grounds the field — at bagasse ash densities a starved ash conveyor can do this within a shift, which is why hopper instruments and the ash-handling system are engineered together. Modern ESP controllers log all of this per field and make energisation trends available to the plant historian; during retrofits we integrate those signals into the boiler DCS so the operators who run the fans also see the collector they are loading.
05 — Measurement basis
Calibration, and trusting the instrument
A number without its basis is an anecdote with decimals.
Two disciplines make the whole set worth its cost. The first is calibration: zirconia O₂ probes checked against certified span gas on a fixed cycle, DP transmitters zeroed at process conditions, opacity monitors given their clear-path zero, and stack thermocouples verified in place — a drifted instrument is worse than none, because it is believed. The second is the "trust or verify" rule from test-review practice: when an instrument reading contradicts expectation, either trust it and act, or verify it within the day — but never let it stand disbelieved and unchecked, because that is how a real low-level or high-CO condition gets normalised as "that transmitter always reads funny".
One honest paragraph on measurement basis, from reviewing before/after performance tests: the commonest way such a test fails is not instrument error but an unrecorded basis. Boiler efficiency by the indirect method depends on feedwater temperature, fuel moisture, ambient temperature and the O₂ at which losses are referenced; if the baseline test did not record feedwater temperature — and a surprising number do not — then a post-retrofit comparison is arithmetic on air. A genuine +3.1–4.0 pp economizer gain can vanish, or double, on paper. Before any upgrade we therefore log the baseline with every input stated: this is standard scope inside an energy audit, and the logged basis is written into the test protocol both parties sign. What Arrow supplies and integrates during retrofits is exactly the set on this page — transmitters, analysers, opacity and temperature instruments and their DCS integration, under our ISO 9001:2015 scope covering spare parts, installation and maintenance services — with the calibration records handed over as part of commissioning.
FAQ
Engineering questions, answered
What is three-element drum level control?
A control scheme using three measurements: drum level, steam flow and feedwater flow. Steam flow acts as feedforward so the feed valve responds to a load change immediately, feed flow closes a fast inner loop, and level trims the balance. It rides through the false level swings, shrink and swell, that defeat single-element control on load changes.
What oxygen level should a biomass boiler run at?
Trim toward an excess-air factor near lambda 1.15, roughly 3 percent O2 dry in the flue gas, with a CO monitor guarding the lower boundary. Biomass units drifting at high excess air give away 0.7 to 0.8 percentage points of efficiency in stack loss, and the extra gas mass loads the fans and the ESP.
Which instruments say the most about boiler health for the least money?
Stack temperature and stack opacity. A stack temperature creeping upward at constant load means fouled heat-transfer surfaces, roughly one percentage point of efficiency lost per 20 degrees C of rise. Opacity rising at constant load flags collector deterioration, a failed ESP field or a bag leak, days before any emission test would.
Why do before-and-after performance tests sometimes prove nothing?
Because the measurement basis was not recorded. Boiler efficiency comparisons fail most often on feedwater temperature: if it was unrecorded in the baseline test, the before and after numbers are not comparable, and a real 3-4 percentage point economizer gain can be masked or doubled on paper. Record every input on both tests.
What ESP instrumentation should a plant monitor continuously?
Per field: secondary voltage in kV and current in mA from each transformer-rectifier set, plus spark rate. Plus stack opacity and hopper level switches. A field whose kV/mA operating point drifts tells you about dust resistivity or internal misalignment, and a full hopper grounding a field is the most common avoidable ESP trip.
Related engineering pages
Send us your plant data
Fuel, boiler capacity, gas flow, current emission and the limit you must meet. An Arrow engineer replies with a technical assessment basis — not a brochure.